A Lightweight Cache-Assisted Data Recovery Mechanism for Latency-Critical Data in IIoT

Latency-critical applications in the industrial Inter-net of Things (IIoT) pose significant challenges to the networks in terms of transmission latency [1]. Especially when packet loss occurs, existing retransmission mechanisms that rely on end-to-end recovery involve long recovery latency, which significantly affects latency-critical applications. To reduce the long recovery latency, we propose a cache-assisted data recovery mechanism that can cache recently transmitted data at the intermediate nodes, e.g., switches and routers. More specifically, we set different timeout period values to different types of data to reduce the data loss detection latency and propose a lightweight replacement policy by considering data latency requirement, data size, and number of receivers to improve the caching efficiency. The simulation results show that the proposed approach has an up to 13 times improvement in terms of unsatisfied data ratio compared to traditional data recovery mechanisms without caching.